Quantifying Privacy Risks of Masked Language Models Using Membership Inference Attacks. (arXiv:2203.03929v2 [cs.LG] UPDATED)

The wide adoption and application of Masked language models~(MLMs) on
sensitive data (from legal to medical) necessitates a thorough quantitative
investigation into their privacy vulnerabilities -- to what extent do MLMs leak
information about their training data? Prior attempts at measuring leakage of
MLMs via membership inference attacks have been inconclusive, implying the
potential robustness of MLMs to privacy attacks. In this work, we posit that
prior attempts were inconclusive because they based their attack solely on the
MLM's model …

attacks language privacy risks